[python-package] add more type hints on Dataset (#5431)

python-package/lightgbm/basic.py

@@ -619,7 +619,10 @@ def _dump_pandas_categorical(pandas_categorical, file_name=None):
     return pandas_str
 
 
-def _load_pandas_categorical(file_name=None, model_str=None):
+def _load_pandas_categorical(
+    file_name: Optional[Union[str, Path]] = None,
+    model_str: Optional[str] = None
+) -> Optional[str]:
     pandas_key = 'pandas_categorical:'
     offset = -len(pandas_key)
     if file_name is not None:
@@ -1879,7 +1882,15 @@ class Dataset:
             self.feature_name = self.get_feature_name()
         return self
 
-    def create_valid(self, data, label=None, weight=None, group=None, init_score=None, params=None):
+    def create_valid(
+        self,
+        data,
+        label=None,
+        weight=None,
+        group=None,
+        init_score=None,
+        params: Optional[Dict[str, Any]] = None
+    ) -> "Dataset":
         """Create validation data align with current Dataset.
 
         Parameters
@@ -1966,7 +1977,7 @@ class Dataset:
             c_str(str(filename))))
         return self
 
-    def _update_params(self, params):
+    def _update_params(self, params: Optional[Dict[str, Any]]) -> "Dataset":
         if not params:
             return self
         params = deepcopy(params)
@@ -1999,7 +2010,11 @@ class Dataset:
             self.params_back_up = None
         return self
 
-    def set_field(self, field_name, data):
+    def set_field(
+        self,
+        field_name: str,
+        data
+    ) -> "Dataset":
         """Set property into the Dataset.
 
         Parameters
@@ -2135,7 +2150,10 @@ class Dataset:
             raise LightGBMError("Cannot set categorical feature after freed raw data, "
                                 "set free_raw_data=False when construct Dataset to avoid this.")
 
-    def _set_predictor(self, predictor):
+    def _set_predictor(
+        self,
+        predictor: Optional[_InnerPredictor]
+    ) -> "Dataset":
         """Set predictor for continued training.
 
         It is not recommended for user to call this function.
@@ -2156,7 +2174,7 @@ class Dataset:
                                 "set free_raw_data=False when construct Dataset to avoid this.")
         return self
 
-    def set_reference(self, reference):
+    def set_reference(self, reference: "Dataset") -> "Dataset":
         """Set reference Dataset.
 
         Parameters
@@ -2207,7 +2225,7 @@ class Dataset:
                 ctypes.c_int(len(feature_name))))
         return self
 
-    def set_label(self, label):
+    def set_label(self, label) -> "Dataset":
         """Set label of Dataset.
 
         Parameters
@@ -2227,7 +2245,7 @@ class Dataset:
             self.label = self.get_field('label')  # original values can be modified at cpp side
         return self
 
-    def set_weight(self, weight):
+    def set_weight(self, weight) -> "Dataset":
         """Set weight of each instance.
 
         Parameters
@@ -2249,7 +2267,7 @@ class Dataset:
             self.weight = self.get_field('weight')  # original values can be modified at cpp side
         return self
 
-    def set_init_score(self, init_score):
+    def set_init_score(self, init_score) -> "Dataset":
         """Set init score of Booster to start from.
 
         Parameters
@@ -2268,7 +2286,7 @@ class Dataset:
             self.init_score = self.get_field('init_score')  # original values can be modified at cpp side
         return self
 
-    def set_group(self, group):
+    def set_group(self, group) -> "Dataset":
         """Set group size of Dataset (used for ranking).
 
         Parameters
@@ -2330,7 +2348,7 @@ class Dataset:
                 ptr_string_buffers))
         return [string_buffers[i].value.decode('utf-8') for i in range(num_feature)]
 
-    def get_label(self):
+    def get_label(self) -> Optional[np.ndarray]:
         """Get the label of the Dataset.
 
         Returns
@@ -2342,7 +2360,7 @@ class Dataset:
             self.label = self.get_field('label')
         return self.label
 
-    def get_weight(self):
+    def get_weight(self) -> Optional[np.ndarray]:
         """Get the weight of the Dataset.
 
         Returns
@@ -2354,7 +2372,7 @@ class Dataset:
             self.weight = self.get_field('weight')
         return self.weight
 
-    def get_init_score(self):
+    def get_init_score(self) -> Optional[np.ndarray]:
         """Get the initial score of the Dataset.
 
         Returns
@@ -2473,7 +2491,7 @@ class Dataset:
         else:
             raise LightGBMError("Cannot get feature_num_bin before construct dataset")
 
-    def get_ref_chain(self, ref_limit=100):
+    def get_ref_chain(self, ref_limit: int = 100) -> Set["Dataset"]:
         """Get a chain of Dataset objects.
 
         Starts with r, then goes to r.reference (if exists),